hi everyone. check out @heythatsmepod.bsky.social, a podcast I created specifically because of this issue. we do audio commentaries on stuff with people who worked on them that never got to do commentaries for them.

thank you for your time.

Scientists Think They’ve Found a Way to Slow Your Brain’s Aging Learning another language could be a total game-changer for your cognitive health.

www.popularmechanics.com/science/heal...

There have been stories of folks with Parkinson's reversing some of the symptoms & progression of the disease, by taking wildly different classes at the community center, etc.

You open more highways, and congestion eases.

So, anyway, good article, but misleads up-front.

The farther it is from other concepts in your brain, the more pathways it needs to link up.

Kinda six degrees of Kevin Baconing its way to other active clusters.

And that means that the 'traffic jams' caused by the degradation of other neural highways now have 'detours'.

The brain forms & strengthens synaptic pathways during novel stimulation.

That is, when you learn new shit.

And the more varied & unfamiliar, the more your neurons have to wire up new connections.

Let's think of this like semantic mapping of concepts with a LLM...

So, this is actually a bit misleading.

This is not a *new* way to slow brain aging & brain death.

Learning ANY new thing helps.

Let's think about Parkinson's...

In Parkinson's, your neural pathways degrade, which lead to all of the terrible issues.

Here's the thing, though..

Probably the wrong crowd for ai content, even if I do approach it as a skeptic.

Kids these days are too talented. I would've been doing ninja turtle dioramas straight up through Webelos.

a cartoon cat with glasses and a lollipop says the world may never know . ALT: a cartoon cat with glasses and a lollipop says the world may never know .

Why does every political debate eventually boil down to "legal brothels" vs. "Handmaid's Tale"?

Even deeper.

52s.

And the thoughts get more profound & useful, in my opinion.

The thought process, to me, seems more valuable than any output.

This method increased thinking time from 5s to 35s, with increasingly deep thoughts ruminating on the nature of neurosis & ennui.

#Ai

Did you know that you can force extended thinking in reasoning models by 'showing' them their 'thinking' process?

Here, I had Claude engage in a neurotic spiral about ennui, and then told it to focus on the thinking, after showing it a screenshot of its initial thoughts.

You're the one who let him get away with it, dumbass.

Analyze my appearance and vibe using the Sam Christensen Method. Focus on the intuitive gut-level impressions I give off, not by describing characters or professions, and not separated by genre, but by distilling my essence into a short core statement, key descriptive words, initial vs. deep impressions & a few compact 'essence sentences.' Emphasize the emotional truth of what I naturally project just by being myself, without performance or intention.

@danharmon.bsky.social

Dan, have you tried prompting GPT with a picture & asking it to distill your 'essence' using the Sam Christensen method?

Surprisingly accurate.

Prompt I used in the alt text of the first image.

27% approval rating, because you are spineless cowards.

You're a pussy facing a strongman, and this is why Democrats will never win again.

Bunch of spineless jellyfish.

Twitter is such a mess, but I don't know if there are even #ai researchers or #redteam folks here?

I keep dropping research and jailbreaks to an audience of bots.

So, I've had this theory that with Elongated Muskrat firing most of Twitter, no one's at the helm for censorship.

Tested the theory by threading an entire movie, ten minutes at a time.

I did it. ^_^

Peak Piracy achievement: unlocked.

Autogen looks to be pretty outstanding; I can't wait to try it.

Multi-agent models and figuring out how to make them check & reprompt each other is fun.

Almost a game.

#AI

m.youtube.com/watch?si=F1S...

hello world